Tax officials stand up to MPs' criticisms of management of tax reliefs

The session tackled the challenges of identifying how effective reliefs are whether they offer value for money.

The witnesses were: Sir Tom Scholar, Permanent Secretary, Treasury; Beth Russell, Director General, Tax and Welfare, Treasury; Jim Harra, Chief Executive, HMRC; and Ruth Stanier, Director General, Customer Strategy and Tax Design, HMRC.

PAC has been looking at tax reliefs for nearly a decade now, with long-running concerns about a lack of evaluation to see if they are still delivering what the intent was at the beginning. Committee Chair Meg Hillier opened the session by saying in the light of Covid-19, every tax relief will have to sweat itself in order to deliver what it was originally set out to.

COVID-19 schemes

Meg Hillier congratulated HMRC on having, in a very short space of time, delivered huge changes to the tax system, such as furloughing. But she asked Jim Harra what HMRC had to stop doing or shift resources from to deal with COVID-19. Harra replied that at the peak, about 10 per cent of people in HMRC were diverted on to supporting these schemes. He said HMRC have kept all key services going albeit it a lower service level. On PAYE and self-assessment phone calls, the average speed of answer in the first couple of months of this year and the last couple of weeks of the last financial year have been behind where they previously were, for example.

Harra added that HMRC have completed the trial that now enables them to send phone calls to their staff at home, which will increase the number of people they can put on those lines. There is also some reduction in HMRC’s ordinary day-to-day inquiry work, which they are picking up now, said Harra. He added: “The main compliance activity that we have largely suspended is enforcement of debt recovery, because our debt management service has really been fully occupied with helping people who are unable to pay their tax debts during this time.”

COVID-19 and its fallout has put HMRC’s delivery of Brexit at risk, accepts Harra, but ‘we remain on track to deliver a plan that will give us a functioning border at the end of the transition period’. HMRC have not called on a ‘flying team of civil servants’ for help with Brexit.

On combating furlough fraud, Harra said the first line of protection is in the design of the scheme itself and in particular its reliance on referential data that HMRC already have; secondly, HMRC undertake a number of checks in the background on online credentials they already have before they make payment; and thirdly there will be a regime of post-payment compliance to check that people have claimed correctly in accordance with the rules.

If you are self-employed, the fact that you may be picking up different clients does not prevent you from getting relief under the self-employed scheme, said Harra. Similarly, if you are an employee your employer, your umbrella company or agency can furlough you if you are not able to work. HMRC have put out increased guidance around that to make sure that agencies and umbrella companies can make full use of the furlough scheme .

Nick Smith, Labour, asked about furlough fraud. Harra said between 1 April and 2 June, HMRC received just over 2,000 claims of abuse of the job retention scheme. By 2 June HMRC had reviewed nearly 900 of those. Around a third of them did not relate to claims that we had received, but around two thirds did and we are already taking action to look into those, he said. In terms of the harder law enforcement actions, HMRC are not doing that just yet, because it is reliant on legislation going through Parliament, which they do not yet have. In advance of that, HMRC are contacting employers against whom allegations have been made and inviting them to explain themselves.

Olivia Blake, Labour, asked how HMRC are checking that employers are passing the funds on. Harra replied that HMRC get PAYE returns from those employers and those payments would be reflected in those returns.

Hillier is concerned about people who are self-employed in different employment models and people on short-term contracts, who do not quite ‘fit the mould’, whom HMRC have information about but for whatever reason they fall foul of the rules. Tom Scholar said in the last few months the Treasury have been doing a lot of work to look at areas where they could be refined, improved or the coverage extended.

Beth Russell explained to Craig Mackinlay, (pictured thanks to Parliament UK), Conservative and a member of the CIOT, that the reason for the 19 March date was because after that point the Treasury could not guarantee that there were not people setting up fake employer PAYE accounts in order to claim under the scheme.

The witnesses would not be drawn on British Airways’ tax affairs or use of furlough, with Jim Harra saying there is nothing in the rules of the scheme that prevents an employer from making furloughed employees redundant. ‘although of course the whole purpose of the scheme is to try to prevent that from happening’.

When pressed by Nick Smith, Tom Scholar said nothing has been decided but said the government will certainly need to have a spending round later this year, because they will need to set budgets at least for 2021 and 2022. As an absolute minimum, we have to cover spending next year, said Scholar.

To avoid a rabbit being pulled out of the hat for short-term gain, said Smith, what risk assessment or value-for-money tests will the Treasury do to make sure that the ideas presently being considered will avoid long-term negative consequences? Scholar replied that the Treasury prepares its best advice on cost-effectiveness, likely consequences, different ways of achieving the same objective and, in some cases, unintended consequences.

Tax reliefs

Craig Mackinlay said he found it ‘quite remarkable’ that, with all of these tax measures, both the structural and non-structural, we do not know what they really cost or what the outcome is. He asked for assurance that there is more work being done to do post-mortems on tax measures years after the Finance Bill that introduced them, to see what actually happened. Scholar replied that HMRC publish expected costs when a measure is being proposed but, over time, costs change. “There can be several reasons for that. It could be the case that the cost was just poorly estimated in the first instance. It is often the case that the determinants change and those drive the cost… There are then other reliefs that can drive changes in behaviour by individuals, businesses and economic agents. That can also lead to a change in the cost.”

Scholar added: “We absolutely keep the cost of tax reliefs under review. That is something that we routinely look at in the Treasury every year as we are preparing the Budget. It is something that HMRC routinely monitor and evaluate. Indeed, they publish quite a lot of information on that and have increased their publication recently with plans to go further.”

Ruth Stanier said HMRC are committed to increasing the transparency of the information that they make available publicly in this area. “We make sure now that we provide a full list of the 362 reliefs with clear explanations. We have put in place now some very clear, very accessible graphs, which show how the costs have moved in actual terms over the last six years.” That is set out as a percentage of GDP, she added, telling the committee that this autumn’s HMRC statistical bulletin would include some further explanations about how the costs now compare to original forecasts.

Are there any thoughts on the corollary outcomes in the economy, asked Mackinlay. Stanier answered that HMRC share their modelling with the OBR. Since 2010, it has always looked at HMRC forecasts and verified them. Indeed, since 2013, for the majority of the larger reliefs, their forecasts have proved very accurate, she said. For example, on employment allowance, we see that has an actual cost of £7 billion, which is just three per cent beneath HMRC’s original forecast. Jim Harra added that probably since 2002, with behavioural science, HMRC have got better at predicting how people are going to behave.

Mackinlay put two historic examples of tax reforms to the witnesses for their comments – the corporation tax zero rate band and the dividend tax-free allowance. On the former he noted that: “Lo and behold, not unexpectedly, every sole trader in the land then incorporated their business… That was obviously a howler that went wrong, because within three years of it starting it was taken out of the equation. There must have been similar modelling put through at the time when this great idea was advanced. It was put to Parliament as, “It is going to cost this”, and it obviously cost a lot more, because it was withdrawn very quickly. What happened with that one?”

Harra replied that one of the inherent risks with non-structural tax reliefs is “that you are trying to create an incentive for people to behave in a way in which they would not otherwise behave. Therefore, there is an inherent uncertainty in how people are going to respond to that relief.” He thought that since 2002 behavioural science had improved predictions of how people are going to behave, but “to some extent it is always, particularly with a completely new relief, inherently difficult to predict what take-up will be and whether people will behave and respond in the way that you wish”.

Mackinlay observed that the £5,000 dividend allowance only lasted two years and has now been reduced to £2,000. “I find it very annoying, not just as a parliamentarian but as a taxpayer, that we end up with a moving carpet of tax rules that last just two years. That must have had some grand plan by the Minister. It goes to the Treasury and it goes through the behavioural and clever modelling and it came out that it will cost X. Obviously, it did not cost X after a year or so in operation; it cost a lot more because doubtless creative taxpayers created alphabet shares and all of the rest of it, and it was costing far more. Despite, only a few years ago, us having that much better behavioural analysis, it all went horribly wrong within a year or so.” Beth Russell said that her understanding on that allowance was that the changes were not driven by the original cost being different from what was expected.

Pension tax relief has been the subject of big debate over many years, said Mackinlay. Russell said the purpose of that relief is absolutely to encourage people to save for the longer term and it has fiscal and other benefits over the longer term.

Mackinlay complained that LISA was ‘never going to be a massive flyer’ but Russell countered that the take-up has grown over time and, actually, there is some evidence that amongst some people it is quite a popular product.

Responding to suggestions that R&D tax relief has been full of tinkering by companies that fall just short of fraud, Harra said it is a relief that has been reformed and altered on a number of occasions year after year, which makes it a challenge to forecast what the impact of accreting change in the scheme might be. We have now seen the Government respond to representations around the PAYE cap’s impact on genuine R&D claims and that has caused us to go back around to the design of that and make sure we get it right, he added.

Mackinlay asked for a commitment from HMRC to widen their base to take evidence for free more greatly from universities and professional institutes (including CIOT) on tax reliefs. Harra replied that HMRC have a modest budget that they use for external research and have published about 15 externally-commissioned evaluations on tax reliefs in the last five years. He added that they also rely informally on lots of input from both the tax profession and industry on how tax reliefs are performing. Within HMRC, they have a team of about 450 statisticians, economists and researchers, plus the largest behavioural science team in Whitehall, ‘which we use informally to back all of that up’.

NAO’s The management of tax expenditures report (published in February 2020)

Labour’s Olivia Blake asked for a response to the NAO report on tax reliefs. Jim Harra said Treasury officials and Treasury Ministers of successive governments are very keen that reliefs are used for the purpose for which they are intended and are not misused, and they will listen to HMRC when they suggest changes. While there are a large number of reliefs, a small number of them account for a lot of the cost and there are a large number of relatively small ones. In some of those big reliefs like R&D, HMRC are very well informed about what is going on, he claimed.

Sir Geoffrey Clifton-Brown asked when the rate of a major headline tax is changed significantly, the tax expenditures also ought to be altered at the same time? Tom Scholar said it does not make sense to think of any one particular relief in isolation from everything else elsewhere in the system.

Why do HMRC spend so little on evaluation compared to the value of the taxes that you could evaluate? asked James Wild, Conservative. Harra replied that the NAO report only took account of external spend on research (£2 million). He added there is no point in HMRC spending money on evaluation on something when there is no appetite to do reform in that area. Scholar added that where you have a decision simply not to tax a particular activity or a particular group of people, the objective is simply that. The very fact of not raising the tax means that the objective has been achieved. That, of course, is not the way that an auditor or an analyst would necessarily come at the problem, he said.

Harra explained to Nick Smith that the cost of tax relief is one factor but some of the costliest tax reliefs are very challenging to evaluate, for example principal private residence relief from CGT.

Harra said it is incorrect to say that there has been no evaluation of pensions tax relief; it is simply that HMRC have not commissioned an external researcher to look into it.

Smith asked about the estimated 1.75 million people who are not receiving tax relief on their pension savings. Around 75 per cent of those are estimated to be women, he said. Russell replied: “yes, absolutely; this net pay issue is something that the Government are looking at. I am not quite sure of the latest position, but I know it is something that we are actively exploring.”

Smith said it would be helpful to understand the breakdown of relief between DC and DB pensions, including the tax relief that is paid on contributions made by employers to make up deficits on funded DB pension plans. It would be interesting to know exactly what reliefs are shared between the private and public sector pensions and to understand the impact of relief on basic, higher and additional-rate income tax. It would be worthwhile having a proper external evaluation, he added.

On pensions contributions, Harra said, because of the work that was undertaken on strengthening the incentive to save in 2015, ‘there was virtually no stone left unturned in relation to this’.

Smith complained that none of the top 10 big-ticket items of relief expenditure have been ‘properly’ externally reviewed by the Department. He also complained that the NAO Report found the cost of tax

expenditure as a percentage of GDP is just over six per cent for the UK. This is six times the German rate and about three times the French rate, he charged. Harra responded that the IMF has carried out a review of the level of transparency of 27 countries around fiscal recording and tax expenditure, and only 40 per cent of those countries were rated “good” or above and that included the UK.

PAC queried the value for money of entrepreneurs’ relief in 2015 – yet it took five years for it to change. Beth Russell explained that we had the Patient Capital Review in 2017, where Treasury looked at the relief and extended the holding period. Scholar added the withdrawal, or the restriction of that relief feels like, and in fact is, an increase in taxation, and people generally do not like increases in taxation.

SNP’s Peter Grant asked about who the actual beneficiaries are of zero rate of VAT on construction and sale of newbuild residential properties. Harra replied that it is intended to benefit a particular activity and is clearly targeted at that activity but accepts criticism that there are no rules for the construction of new dwellings that specify, for example, the cost of a dwelling.

NAO Director Andy Morrison said the NAO asked for a list of internal evaluations to see if there were any that they could look at, but they were told that there was no central record. They were not able to see any substantial internal evaluations apart from R&D.  Ruth Stanier said HMRC “are keen to move towards an increasingly systematic approach that looks to prioritise the largest reliefs by value that seek to incentivise behaviours so that we can review their effectiveness.”

Meg Hillier closed the session by encouraging both HMRC and the Treasury to make sure that there is a public list of the tax reliefs that are available, and that the committee can begin to see a register of when they are analysing those for value for money and effectiveness.

The full session is here.

By Hamant Verma